I've run into two problems writing functions for data frames. I often get a data frames with 2 variables, and I want to recode them into one variable.
If V1>0 and V2 <0 then new_variable = "V1>0, V2<0.
In all of dataframes I've got V1 and V2 have different names.
Problem number 1. I don't know why test_df$newVar, after this function get only "C>0, I>0"
#Using test FUN on example data frame
test_df.afterFUN <- test_fun(test_df, var1 = "V1", var2 = "V2", newVar = "category")
Problem number 2. Why last argument of this function "newVar" doesn't change name into "category"? If I run the code of this function fitted to single data frame (renaming variable and ect.) it will work and give me what I want (look test_df2)
rm(list = ls())
library("dplyr") # for filter
# Preparing example data frame
rama <- rbind(c(-5:20, -20:5), c(-20:5, -5:20))
rama <- t(rama)
colnames(rama) <- c("V1", "V2")
test_df <- as.data.frame(rama)
#Test FUN
test_fun <- function(df, var1, var2, newVar) {
df1 <- filter(df, var1 == 0, var2 == 0)
df1 <- mutate(df1, newVar = "C=0, I=0")
df2 <- filter(df, var1 == 0, var2 > 0)
df2 <- mutate(df2, newVar = "C=0, I>0")
df3 <- filter(df, var1 == 0, var2 < 0)
df3 <- mutate(df3, newVar = "C=0, I<0")
df4 <- filter(df, var1 > 0, var2 == 0)
df4 <- mutate(df4, newVar = "C>0, I=0")
df5 <- filter(df, var1 > 0, var2 > 0)
df5 <- mutate(df5, newVar = "C>0, I>0")
df6 <- filter(df, var1 > 0, var2 < 0)
df6 <- mutate(df6, newVar = "C>0, I<0")
df7 <- filter(df, var1 < 0, var2 == 0)
df7 <- mutate(df7, newVar = "C<0, I=0")
df8 <- filter(df, var1 < 0, var2 > 0)
df8 <- mutate(df8, newVar = "C<0, I>0")
df9 <- filter(df, var1 < 0, var2 < 0)
df9 <- mutate(df9, newVar = "C<0, I<0")
df <- rbind(df1, df2, df3, df4, df5, df6, df7, df8, df9)
return(df)
}
#Using test FUN on example data frame
test_df.afterFUN <- test_fun(test_df, var1 = "V1", var2 = "V2", newVar = "category")
# Procedure outside of funcion fitted to test_df
df1 <- filter(test_df, V1 == 0, V2 == 0)
df1 <- mutate(df1, newVar = "C=0, I=0")
df2 <- filter(test_df, V1 == 0, V2 > 0)
df2 <- mutate(df2, newVar = "C=0, I>0")
df3 <- filter(test_df, V1 == 0, V2 < 0)
df3 <- mutate(df3, newVar = "C=0, I<0")
df4 <- filter(test_df, V1 > 0, V2 == 0)
df4 <- mutate(df4, newVar = "C>0, I=0")
df5 <- filter(test_df, V1 > 0, V2 > 0)
df5 <- mutate(df5, newVar = "C>0, I>0")
df6 <- filter(test_df, V1 > 0, V2 < 0)
df6 <- mutate(df6, newVar = "C>0, I<0")
df7 <- filter(test_df, V1 < 0, V2 == 0)
df7 <- mutate(df7, newVar = "C<0, I=0")
df8 <- filter(test_df, V1 < 0, V2 > 0)
df8 <- mutate(df8, newVar = "C<0, I>0")
df9 <- filter(test_df, V1 < 0, V2 < 0)
df9 <- mutate(df9, newVar = "C<0, I<0")
test_df2 <- rbind(df1, df2, df3, df4, df5, df6, df7, df8, df9)
This can probably be written nicer, but try:
test_fun <- function(df,col1, col2, newVar) {
temp <- sapply(df[,c(col1,col2)],function(x) revalue(factor(sign(x)),c("-1"="<0","0"="=0","1"=">0")))
df[,newVar] <- apply(temp, 1, function(y) paste0(col1,y[1],", ",col2,y[2]))
df
}
head(test_fun(test_df,"V1", "V2", "category"))
# V1 V2 category
# 1 -5 -20 V1<0, V2<0
# 2 -4 -19 V1<0, V2<0
# 3 -3 -18 V1<0, V2<0
# 4 -2 -17 V1<0, V2<0
# 5 -1 -16 V1<0, V2<0
# 6 0 -15 V1=0, V2<0
Explanation
We use sign
to get the sign of each number within a column (returns -1, 0 or 1). We then rewrite those numbers as the strings "<0","=0" and ">0" using revalue(factor),c())
. We use sapply
to apply it to both columns of test_df
. This returns a character matrix. We then apply paste
to each row to get the character vector that you want. Finally, we assign that vector to test_df$category
.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.